# How LLMs Fail to Support Fact-Checking

arXiv: https://arxiv.org/abs/2503.01902 · HTML: https://mlsanity.com/paper/2503.01902/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 9 of its 18 references that are in the map.
- First posted on arXiv 28 Feb 2025.

## Builds on (most cited first, 9 of 9)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Constitutional AI: Harmlessness from AI Feedback](https://mlsanity.com/paper/2212.08073/) (2022)
- [Whose Opinions Do Language Models Reflect?](https://mlsanity.com/paper/2303.17548/) (2023)
- [From Pretraining Data to Language Models to Downstream Tasks: Tracking the Trails of Political Biases Leading to Unfair NLP Models](https://mlsanity.com/paper/2305.08283/) (2023)
- [Factuality Challenges in the Era of Large Language Models](https://mlsanity.com/paper/2310.05189/) (2023)
- [Large Language Models Produce Responses Perceived to be Empathic](https://mlsanity.com/paper/2403.18148/) (2024)
- [Fact-checking information from large language models can decrease headline discernment](https://mlsanity.com/paper/2308.10800/) (2023)
- [FakeGPT: Fake News Generation, Explanation and Detection of Large Language Models](https://mlsanity.com/paper/2310.05046/) (2023)

## Built on by (most cited first, 2 of 2)

- [Replicating Human Motivated Reasoning Studies with LLMs](https://mlsanity.com/paper/2601.16130/) (2026)
- [Multi-Agent Retrieval-Augmented Framework for Evidence-Based Counterspeech Against Health Misinformation](https://mlsanity.com/paper/2507.07307/) (2025)

## Abstract (arXiv)

While Large Language Models (LLMs) can amplify online misinformation, they also show promise in tackling misinformation. In this paper, we empirically study the capabilities of three LLMs -- ChatGPT, Gemini, and Claude -- in countering political misinformation. We implement a two-step, chain-of-thought prompting approach, where models first identify credible sources for a given claim and then generate persuasive responses. Our findings suggest that models struggle to ground their responses in real news sources, and tend to prefer citing left-leaning sources. We also observe varying degrees of response diversity among models. Our findings highlight concerns about using LLMs for fact-checking through only prompt-engineering, emphasizing the need for more robust guardrails. Our results have implications for both researchers and non-technical users.